JPH09167246A - Method for processing picture and its device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically execute judgement without designating a table attribute such as whether a manually inputted table picture has a table with a horizontal ruled line or not by judging the attribute of picture information in accordance with the characteristic of a row picture which is picked-up from picture information and storing the judged attribute in accordance with picture information. SOLUTION: The attribute of picture information is judged in accordance with the characteristic of the row picture which is picked-up from picture information and the judged attribute is stored in accordance with picture information. That is, in a picture processing device, picture information is inputted from an input part 101, the area of the input picture is divided by an area dividing part 106 and a picture data area such as sentence, separator, figure and table, etc. Information of the respective picked-up picture data areas is stored in a storage part 103. When the table picture is added in the input picture in the result of area division, a processing for automatically discriminating the table picture is executed. In judgement result, the table processing of the table without the horizontal ruled line is executed when the table picture is the table without the horizontal ruled line and the table processing of the table with the horizontal ruled line is executed at the time of the horizontal ruled line.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus such as an OCR (optical character recognition) apparatus, a copying machine, a facsimile machine, etc., which is capable of properly extracting characters from an input image including a front image. The present invention relates to a method and an apparatus.

The present invention relates to an image processing method and apparatus capable of appropriately extracting a character image by judging the format of the input table when extracting the character image from the input table image.

According to the present invention, the user does not manually specify the attributes of the front image included in the input image.
The present invention relates to an image processing method and apparatus which can be automatically judged.

[0004]

2. Description of the Related Art In an OCR device having an automatic area dividing function, an image surrounded by ruled lines is automatically discriminated as a table image, and a cell surrounded by the ruled lines of the table is extracted as one character area. Is going.

After the character recognition, a tab code is inserted between the character strings of the recognition result obtained from one character area,
By inserting a return code between lines, the cell information extracted from the table image is reflected in the document data after character recognition so that the data can be easily poured into the spreadsheet software or the like.

[0006]

However, in the above-mentioned conventional technique, when the table image to be character-recognized is a table without horizontal ruled lines, as shown in FIG. Since it is judged as one character area, a document different from the input table form will be obtained, which will force the operator to arrange the form into the input table form, increasing the burden on the operator. Was there.

[0007]

In order to solve the above-mentioned problems, the present invention extracts a line image from image information, determines the attribute of the image information according to the characteristics of the extracted line image, and makes the determination. The attribute is stored in association with the image information.

In order to solve the above-mentioned problems, the image information according to the present invention is image information surrounded by ruled lines.

In order to solve the above problems, in the present invention, the image information is optically read image information.

In order to solve the above problems, the present invention character-recognizes the line image, and outputs a line delimiter code together with a result of the character recognition.

In order to solve the above problems, the present invention displays a document according to the output information.

In order to solve the above problems, the present invention divides input image information into small image regions surrounded by ruled lines,
A row image is extracted from the image information of the divided small image area, and the row image included in the input image information is compared with the positional relationship of the small image area in which the row image is located, and in the small image area. Order by position.

In order to solve the above problems, in the present invention, the input image information is optically read image information.

In order to solve the above problems, the present invention character-recognizes the line image, and outputs a line delimiter code together with the result of the character recognition.

In order to solve the above problems, the present invention determines the attribute of the image information, and determines whether to extract the line image according to the determination of the attribute determining means.

In order to solve the above problems, the present invention displays a document according to the output information.

[0017]

1 is a block diagram showing the functional arrangement of an example of an image processing apparatus according to the present invention. An input unit 101 is for inputting image information to be a target of character recognition and various instructions given by an operator. It also has a keyboard for editing documents. Here, the image information may be input by a scanner that optically reads a document, or by inputting image information that has already been optically read through a storage device or communication means. Reference numeral 102 denotes an arithmetic processing unit (CP) that controls the device and performs arithmetic / control of each processing unit according to a control program of a flowchart described later.
U is a U), and 103 is a storage unit that stores a control program for the arithmetic processing unit 102 to execute a flowchart described later, data used for various processes, or data generated during various processes. An output unit 104 outputs a calculation processing result, an image processing result, and image data.
It includes a display means such as T or a liquid crystal display, or an output terminal for outputting to another terminal or external storage means. Reference numeral 105 denotes a paragraph of a sentence included in the image information, an ordering unit that orders headings, 106 is a region dividing unit that performs a region dividing process of the image information, and 107 is a table image attribute included in the image information. A table attribute automatic judging unit for automatically judging by detecting and analyzing the characteristics of the table image, a table processing unit 108 for extracting the form of the table from the table image included in the image information, and performing a process for documenting it. A character recognition unit 109 extracts a character image from the image information and further recognizes the extracted character image.

FIG. 2 is a flow chart showing the overall processing flow from inputting an image from the input unit 101 to character recognition and outputting as document data from the output unit 104. The control program is stored in the storage unit 103. It is stored and executed under the control of the arithmetic processing unit 102.

Image information is input from the input unit 101 (S
201). The input image information may be displayed by the display unit included in the output unit 104 to be confirmed by the operator. Next, the area dividing unit 106 divides the input image into areas, and extracts image data areas such as sentences, separators, figures, and tables (S202). Information on each of the extracted image data areas is stored in the storage unit 103. For a detailed method of area division, there is a technique disclosed in Japanese Patent Application No. 4-243252. As a result of the area division, if the input image includes the front image (S203YE
In S204, the process proceeds to S204 in which the attribute of the front image is automatically determined (S204). The automatic determination process of the attribute of the front image in S204 is shown in detail in the flowchart of FIG. 3, and will be described in detail later. As a result of the determination of the table attribute in S204, if the table image is a table without horizontal ruled lines (YES in S205), the process proceeds to S206 to perform table processing of the table without horizontal ruled lines. The table processing of the table without horizontal ruled lines in S206 is a process of extracting character units (corresponding to cells) and ordering the extracted character units so as to document the table form of the table image without horizontal ruled lines. However, it will be described later in detail according to the flowchart of FIG. As a result of the determination of the table attribute in S204, if the table image is a table with horizontal ruled lines (NO in S205), the process proceeds to S209 to perform table processing of the table with horizontal ruled lines. The table process of the table with horizontal ruled lines in S209 is a process of extracting cells for ordering and documenting the table form and ordering the extracted cells.
The numbers are given from left to right and from top to bottom. The order is S in the flowchart of FIG. 4 described later.
It is the same as the order in 402.

When it is determined in S203 that there is no table, and when the processing of S206 or S209 is completed, the process proceeds to S207, where the ordering unit 105 orders headings and text paragraphs consisting of sentence parts excluding the table image of the input image. Attach.
This ordering may be performed from right to left when the text is written vertically, and similarly from top to bottom in the vertical direction.

In S208, the character recognition processing unit 109 performs character recognition processing, and S206, S209, and S20.
The character string is output according to the order given in 7, and the documenting of the input image is completed. However, regarding the output of the characters included in the table image, insert a tab code that is a code that indicates a cell delimiter between cells and insert a return code that is a code that indicates a line delimiter between lines. To do.

As described above, by performing the processing shown in the flow chart of FIG. 2, even when the input image includes a table without horizontal ruled lines, the operator does not manually specify that and the horizontal direction is automatically specified. A table without ruled lines can be discriminated and the table form can be retained and documented for the table.

The table attribute automatic determination processing executed by the table attribute automatic determination unit 107 in S204 will be described below with reference to FIG.
This will be described according to the flowchart of FIG. The target of this processing is the image of the area determined to be the front image by the area dividing processing in S202, as illustrated in FIG.

First, cells surrounded by ruled lines of the table image 50 are extracted (cell 1 to cell 4), and the number of character lines in each cell is counted (S301). For example, cells 1 to 3 have 6 rows and cell 4 has 2 rows. Next S302 and S303
The determination process of is performed for all the cells extracted in S301. The process of S302 is a judgment as to whether or not the number of lines counted in S301 is three or more, and the process of S303 is the height a of the character image included in the cell and the interval between the character images in the vertical character direction. for determining the relationship with b,
It is determined whether or not> 3a. S302 or S303
If there is a cell that is determined to be YES, the table image is determined to be a table without horizontal ruled lines, and the attribute of the table without horizontal ruled lines is added to the area information of the table image stored in the storage unit 103. It is stored (S304). If NO is determined in all the cells GS302 and S303 of the target table image, it is determined that the table image has a horizontal ruled line,
In addition to the area information of the table image stored in the storage unit 103, the attributes of the table with horizontal ruled lines are stored. In the table image illustrated in FIG. 5, cells 1 to 3 are determined to have three or more rows in S302, and cell 4 is determined to have a large gap between two rows and b> 3a. The attributes of the ruled table are stored in the storage unit 103.

By the processing of the flow chart of FIG. 3,
It is possible to automatically determine whether the front image included in the input image has horizontal ruled lines or not, without designating manually by an operator.

The horizontal ruled line-less table processing executed by the table processing unit 108 in S206 will be described below with reference to the flowchart of FIG. The target of this processing is the image of the area determined to be the table image by the area dividing processing in S202, such as the table image 60 illustrated in FIG.

First, cells 1 to 1 extracted from the front image 60
Each cell 3 is divided into row unit blocks (S40).
2). Cell 1 is divided into 5 row blocks. Cell 2
And cell 3 is similarly divided into 5 row blocks. S
The blocks extracted in 401 are ordered (S4).
02). Since the ordering is from left to right and from top to bottom, the top block of cell 1 is the first block in the order, and the top block of cell 2 is the top block of cell 3. The block of No. ends in cell 3 in the horizontal direction, so move downward and the second from the top of cell 1,
The cell 2 is ordered from the second from the top. This order information is stored in the storage unit 103 together with the block information.

As described above, since a plurality of items in one cell are divided by extracting blocks in row units and ordering is also performed, a table without horizontal ruled lines has the same character string for each cell as the table with horizontal ruled lines. Extraction and ordering, and further documenting according to that order, preserves table morphology.

[0029]

As described above, according to the present invention,
A line image is extracted from image information, the attribute of the image information is determined according to the characteristics of the extracted line image, and the determined attribute is stored in association with the image information, so that the operator manually inputs it. It is possible to automatically determine the table image without specifying a table attribute such as whether the table image is a table with horizontal ruled lines or not, and it is possible to improve operability.

As described above, according to the present invention, the image information is image information surrounded by ruled lines so that a plurality of items included in one cell surrounded by the ruled lines can be automatically determined. Can be done.

As described above, according to the present invention, the image information is optically read image information, so that the optically read image is automatically read without operating a key or the like. Can be processed.

As described above, according to the present invention, the line image is character-recognized, and the line-separation code is output together with the result of the character recognition, so that the document can be displayed even after the character-recognition and documentation. The shape of can be maintained.

As described above, according to the present invention, by displaying a document according to the output information,
The result of character recognition of the input image and the form of the table can be confirmed on the display screen.

As described above, according to the present invention, input image information is divided into small image areas surrounded by ruled lines,
A row image is extracted from the image information of the divided small image area, and the row image included in the input image information is compared with the positional relationship of the small image area in which the row image is located, and in the small image area. By ordering according to position, it is possible to properly determine the shape of a special table without horizontal ruled lines.

As described above, according to the present invention, the attribute of the image information is determined, and whether or not the line image is extracted is determined according to the determination of the attribute determination means, so that it is appropriate. Since the proper processing is performed only on the table, the form of the table can be maintained even after the document is recognized and recognized.

[Brief description of the drawings]

FIG. 1 is a functional configuration block diagram of an image processing apparatus.

[Fig. 2] Flow chart of overall processing

FIG. 3 is a flowchart of table attribute determination processing.

FIG. 4 is a flowchart of table processing without horizontal ruled lines.

FIG. 5 is an exemplary diagram of a table image for performing table attribute determination processing.

FIG. 6 is an explanatory diagram of table processing without horizontal ruled lines.

FIG. 7 is an explanatory diagram of conventional front image processing.

Claims (20)

[Claims] 1. A line image extracting means for extracting a line image from image information, an attribute determining means for determining an attribute of the image information according to a characteristic of the extracted line image, and the determined attribute as the image information. An image processing device comprising: a storage unit that stores the data in association with each other. 2. The image processing apparatus according to claim 1, wherein the image information is image information surrounded by ruled lines. 3. The image processing apparatus according to claim 1, wherein the image information is optically read image information. 4. The image according to claim 1, further comprising: a character recognition unit for recognizing the line image as a character, and an output unit for outputting a line delimiter code together with a result of character recognition by the character recognition unit. Processing equipment. 5. The image processing apparatus according to claim 4, further comprising display means for displaying a document in accordance with the information output by the output means. 6. An image information input means for inputting image information, an image dividing means for dividing the input image information into small image areas surrounded by ruled lines, and a line image based on the image information of the divided small image areas. A row image extracting means for extracting the row image, and a sequence means for ordering the row images included in the input image information according to the positional relationship of the small image area in which the row image is located and the position in the small image area. An image processing apparatus having. 7. The image processing apparatus according to claim 6, wherein the image information input by the image information input means is optically read image information. 8. The image according to claim 6, further comprising: a character recognizing unit that character-recognizes the line image, and an output unit that outputs a line delimiter code together with a result of character recognition by the character recognizing unit. Processing equipment. 9. An attribute determining unit for determining an attribute of the image information, wherein the line image extracting unit determines whether to extract a line image according to the determination of the attribute determining unit. The image processing device according to claim 6. 10. The image processing apparatus according to claim 9, further comprising display means for displaying a document in accordance with the information output from the output means. 11. A line image is extracted from image information, an attribute of the image information is determined according to a characteristic of the extracted line image, and the determined attribute is stored in association with the image information. Image processing method. 12. The image processing method according to claim 11, wherein the image information is image information surrounded by ruled lines. 13. The image processing method according to claim 11, wherein the image information is optically read image information. 14. The image processing method according to claim 11, wherein character recognition is performed on the line image, and a line delimiter code is output together with a result of the character recognition. 15. The image processing method according to claim 14, wherein the document is displayed according to the output information. 16. The input image information is divided into small image regions surrounded by ruled lines, a line image is extracted from the image information of the divided small image regions, and the line images included in the input image information are extracted. An image processing method characterized in that the row images are ordered in accordance with the positional relationship between the small image areas and the positions in the small image areas. 17. The image processing method according to claim 16, wherein the input image information is image information optically read. 18. The image processing method according to claim 16, wherein character recognition is performed on the line image, and a line delimiter code is output together with a result of the character recognition. 19. The image processing method according to claim 16, wherein the attribute of the image information is determined, and whether to extract the line image is determined according to the determination of the attribute determining unit. 20. The image processing method according to claim 19, wherein a document is displayed according to the output information.